The use of queueing network models was stimulated by the appearance (1975) of the exact product form solution of a class of open, closed and mixed queueing networks obeying the local balance principle and solved, a few years later, by the popular mean value analysis algorithm (1980). Since then, research efforts have been produced to approximate solutions for non-exponential services and non-pure random mechanisms in customer processing and routing. The purpose of this paper is to examine the suitability of modeling choices and solution approaches consolidated in other domains with respect to two key logistic processes in container terminals.,In particular, the analytical solution of queueing networks is assessed for the vessel arrival-departure process and the container internal transfer process with respect to a real terminal of pure transshipment.,Numerical experiments show the extent to which a decomposition-based approximation, under fixed or state-dependent arrival rates, may be suitable for the approximate analysis of the queueing network models.,The limitation of adopting exponential service time distributions and Poisson flows is highlighted.,Comparisons with a simulation-based solution deliver numerical evidence on the companion use of simulation in the daily practice of managing operations in a finite-time horizon under complex policies.,Discussion of some open modeling issues and encouraging results provide some guidelines on future research efforts and/or suitable adaption to container terminal logistics of the large body of techniques and algorithms available nowadays for supporting long-run decisions.

中文翻译：

---

港口物流运作建模的排队分析

遵循局部平衡原理的一类开放，封闭和混合排队网络的确切产品形式解决方案的出现（1975年），刺激了排队网络模型的使用，并在几年后通过流行的均值分析解决了该问题。算法（1980）。从那时起，人们开始进行研究工作以近似解决客户处理和路由中非指数服务和非纯随机机制的解决方案。本文的目的是针对集装箱码头中的两个关键物流流程，检查在其他领域中合并的建模选择和解决方案的适用性。特别是，